Irving Fisher was the first economist to posit what has come to be known as uncovered interest parity relation. He was also the first to offer a peso-problem type explanation for important episodes in which it was violated. After reviewing his theoretical and empirical work on this subject, we go on to reexamine both his data and several other bodies – two consisting of multi-country panel data and a third consisting of a two-century long time series for the United States and the United Kingdom. We find evidence in these data of the important role played by episodic phenomena in disturbing the UIP relation. Like Fisher, we find that the influence of such phenomena dissipates over time. Over the long term these errors prove to be less important and UIP receives empirical support.